Cancer cells are under persistent oxidative stress. Oncogenic transformation such as with Kras, and metabolic alterations result in increased oxidative stress in tumor cells. Tumor cells adapt to persistent oxidative stress by activating redox sensitive transcription factors that increase the expression of endogenous antioxidants, promote survival pathways, induce chemoresistance, and reduce caspase activation. More significantly, ROS also plays an important role in the survival of cancer stem cells. A subset population of cancer stem cells contains lower ROS levels, thus providing protection against DNA damage such as radiotherapy. Moreover, the self-renewal capacity of cancer stem cells is sensitive to cellular ROS levels. Both bulk tumor and cancer stem cells are vulnerable to excess levels of ROS and this characteristic can be exploited for therapy. Our overarching hypothesis is that compounds able to effectively increase the levels of ROS in cancer cells will tip the balance towards cell death and can potentially overcome drug resistance. Recently, we screened a library of highly diverse compounds on an Extracellular Flux Analyzer that measures cellular respiration. Among hundreds of compounds screened, we identified DFC232, a compound that caused a maximum oxygen consumption rate (OCR) in Mia PaCa-2 cells. DFC232 induced rapid onset of ROS production and activation of AKT, followed by a substantial increase in the phosphorylation of the transcription factor FOXO3a, culminating in cell death. DFC232 shows single agent activity in a Mia PaCa-2 xenograft model with no signs of toxicity. In subsequent mechanistic studies, using a novel next-generation sequencing technology (Bru-Seq), we observed that DFC232 produced a remarkable inactivation of mitochondrial gene transcription by potentially affecting the D-loop. Our first round of ADMET-guided lead optimization campaign generated compounds (e.g. DFC325) with nanomolar potency in a panel of PDAC cell lines and remarkable single agent efficacy in mice. Our central hypothesis is that DFC232 and analogs induce ROS production, tipping the balance toward apoptosis. We further hypothesize that DFCs act through a novel mechanism by effectively disrupting transcription from the mitochondrial D-loop. Moreover, DFC analogs are novel agents with unique targets and have biodistribution, safety, and efficacy characteristics necessary for potential clinical benefit in PDAC treatment. To test our hypothesis we will focus on the following specific aims: Aim 1: To perform ADMET, metabolic, and PK-guided synthesis of novel analogs to enhance potency and efficacy. Aim 2: To perform mechanistic studies of top 5 compounds as single agent and in combination with gemcitabine and abraxane using Bru-Seq technology. Aim 3: To determine the in vivo efficacy of top 5 compounds as single agents and in combination with gem/abraxane in orthotopic and genetically engineered mouse models (GEMM) of KRAS driven pancreatic cancer.